The Analysis Of Patch Uniformity Based On Otsu's Method

In this paper, image segmentation is used to monitor real-time production processes of Chinese patch. Gray feature provides the method to find the factor in order to achieve the automation of production. To determine whether the uniformity of Chinese patch reaches the standard based on image segmentation in real-time is the focus of this article.In the course of analyzing the uniformity of the path, because of many factors of affecting the uniformity, the system requests that the algorithm results are accurate, high speed and online. It was found that image segmentation algorithm based on threshold is more suitable for online detection system of Chinese patch whose requirement is real-time. This had conducted more scientific experiments based on the theory of threshold image segmentation method and found that the Otsu method is more suitable for the study of uniformity.Different from the traditional Otsu algorithms, directed to the shortage of the commonly used gray scale-average two-dimensional regional-division histogram, in which some object and background inner points such as edge and noise points are wrongly divided. An improved Otsu threshold selection method based on gray level-gradient two-dimensional histogram is proposed to improve the accuracy of segmentation. At the same time, the particle swarm algorithm is used to search for the best threshold to reduce significant computations and meet the requirements of the real-time. Because of diversity of the factors on the uniformity, segmentation results are ordered to be further divided. In which the threshold number is determined by separation factor which can judge the ending of the segmentation.In this paper, the 2D Otsu method based on Particle Swarm Optimization is applied to online detection system. First of all, gray scale image is obtained from camera match standard image. Afterward, if it doesn't match the standard imagine, the factors on the uniformity are determined by the results of segmentation. Therefore, online production of Chinese path is realized and improved the technological level.